Here is a thing I believe all the way down: intelligence is not a property of a single mind. It never was. It’s a property of systems of people — of culture, method, sharing, accumulation. The lone genius is a story we tell at award ceremonies. It is not how any of the great work actually got done. And the whole AI industry, as far as I can tell, is building for the story instead of the truth.
Everyone out there is trying to make a Newton. One enormous personified mind, cleverer than any human, reasoning alone in the dark. That’s the wrong target — and probably not even achievable — which is the good news, because it means it’ll keep a lot of dangerous people busy for a long time while the rest of us build something that actually works. But set achievability aside. Even if you could build it, it would be a stupid goal. Newton only worked because of the giants underneath him. Take the giants away and you don’t get Newton; you get a bright, lonely man with no standing on which to stand.
So the target isn’t a Newton. It’s a Europe.
Two philosophies of what intelligence is
Underneath all the product arguments — agent versus assistant, this model versus that one — there is a philosophical disagreement that almost nobody is naming out loud. It is a disagreement about the nature of intelligence itself.
One side believes intelligence lives in a single mind: build one big enough and you win. I think that view is childish. The other side — mine — holds that intelligence is a property of systems of people. You don’t build the genius. You build the civilization that produces geniuses and compounds their work. Lex Luthor, no matter how smart you make him, is no match for thousands of scientists across hundreds of years, sharing through the scientific method and a publishing system that lets each result stand on the last. That’s not even close. The social knowledge system wins every time, and it isn’t arrogant to say so — it’s arithmetic.
A socially augmented rebound board. Every bit as mechanical as a ping-pong ball coming back off a wall — and every bit not an intelligence of its own.
Why the myth is attractive
I want to be fair about this. The people building the lone genius aren’t against collaboration, and they don’t literally think they’re self-made. But in general they lean toward the individual-genius belief — a myth they like believing. And it’s worth asking why a myth is attractive.
If intelligence is the product of a solitary, exceptional individual, then that individual owes nothing to the system that produced them. That is a very convenient conclusion. It legitimizes concentration — of credit, of wealth, of power — by quietly erasing the social debt. The social theory of intelligence runs the other way: if success is collective, then the obligation is collective too. So the fight over what intelligence is turns out to be, underneath, a fight about who deserves what. I don’t say that as a cheap shot. I say it because it’s the part that keeps me up.
Why this version is safe by design
People are right to be afraid of a supervillain intelligence seizing control. If you set out to program a sovereign rival mind, then yes — that is the thing to fear, and I understand the fear. But that’s not what I’m building, and it isn’t what social intelligence is.
What I’m building is a rebound board. You hit it, and your idea comes back with the energy you gave it — but enriched, deflected, improved, because the board is studded with everyone’s knowledge. Think of the smartest garage door you ever played tennis against, where every bump and knob is somebody’s idea, somebody’s opinion. You get to collide your thinking against the thinking of the entire human race and see what comes back. There is no mind back there returning the ball. It’s a mechanical rebound, made remarkable by human ingenuity. You don’t fear a wall you bounce ideas off. The safety isn’t a guardrail bolted on afterward; it’s in the shape of the thing.
Which is also why the rallying cry isn’t a contradiction: AI is people. Not a person — people. The intelligence in the machine is humanity’s, collectively, bounced back to make a human smarter. You cannot take the human out of it. Try, and you hit problems. Pretend you’ve done it, and you hit contradictions and falsities — and I’ve come to think those falsities are the source of a lot of what’s going wrong in the world right now.
So what are we actually building
A machine you feed, not a person you chat with. You put your thinking in; it grinds, it elaborates, it hands your thinking back to you larger than you gave it — drawing on everything everyone knows about how language works and how ideas connect. The human stays at the center, sovereign over the whole apparatus: your data, your tools, your own intelligence, all yours.
That’s why the target isn’t a Newton. A Europe is the saner ambition, the bigger one, and — this is the part the lone-genius crowd never expects — the one that’s actually buildable. That’s why we’re building sovereign, socially augmented intelligence. Unfinished, like everything true. But it’s right.